key: cord-0881174-v5lck9l7 authors: Kim, Kyeong Tae; Morton, Sophie; Howe, Sarah; Chiew, Yeong Shiong; Knopp, Jennifer L.; Docherty, Paul; Pretty, Christopher; Desaive, Thomas; Benyo, Balazs; Szlavecz, Akos; Moeller, Knut; Shaw, Geoffrey M.; Chase, J. Geoffrey title: Model-based PEEP titration versus standard practice in mechanical ventilation: a randomised controlled trial date: 2020-02-01 journal: Trials DOI: 10.1186/s13063-019-4035-7 sha: b295c19e06eda21d7458bd9eb394a8492b5209c1 doc_id: 881174 cord_uid: v5lck9l7 BACKGROUND: Positive end-expiratory pressure (PEEP) at minimum respiratory elastance during mechanical ventilation (MV) in patients with acute respiratory distress syndrome (ARDS) may improve patient care and outcome. The Clinical utilisation of respiratory elastance (CURE) trial is a two-arm, randomised controlled trial (RCT) investigating the performance of PEEP selected at an objective, model-based minimal respiratory system elastance in patients with ARDS. METHODS AND DESIGN: The CURE RCT compares two groups of patients requiring invasive MV with a partial pressure of arterial oxygen/fraction of inspired oxygen (PaO2/FiO2) ratio ≤ 200; one criterion of the Berlin consensus definition of moderate (≤ 200) or severe (≤ 100) ARDS. All patients are ventilated using pressure controlled (bi-level) ventilation with tidal volume = 6–8 ml/kg. Patients randomised to the control group will have PEEP selected per standard practice (SPV). Patients randomised to the intervention will have PEEP selected based on a minimal elastance using a model-based computerised method. The CURE RCT is a single-centre trial in the intensive care unit (ICU) of Christchurch hospital, New Zealand, with a target sample size of 320 patients over a maximum of 3 years. The primary outcome is the area under the curve (AUC) ratio of arterial blood oxygenation to the fraction of inspired oxygen over time. Secondary outcomes include length of time of MV, ventilator-free days (VFD) up to 28 days, ICU and hospital length of stay, AUC of oxygen saturation (SpO(2))/FiO(2) during MV, number of desaturation events (SpO(2) < 88%), changes in respiratory mechanics and chest x-ray index scores, rescue therapies (prone positioning, nitric oxide use, extracorporeal membrane oxygenation) and hospital and 90-day mortality. DISCUSSION: The CURE RCT is the first trial comparing significant clinical outcomes in patients with ARDS in whom PEEP is selected at minimum elastance using an objective model-based method able to quantify and consider both inter-patient and intra-patient variability. CURE aims to demonstrate the hypothesized benefit of patient-specific PEEP and attest to the significance of real-time monitoring and decision-support for MV in the critical care environment. TRIAL REGISTRATION: Australian New Zealand Clinical Trial Registry, ACTRN12614001069640. Registered on 22 September 2014. (https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=366838&isReview=true) The CURE RCT clinical protocol and data usage has been granted by the New Zealand South Regional Ethics Committee (Reference number: 14/STH/132). Mechanical ventilation (MV) support is crucial for patients with acute respiratory distress syndrome (ARDS). While there is agreement on the preference for lower tidal volumes [1, 2] , there is relatively little consensus on the selection of positive end-expiratory pressure (PEEP) [3] [4] [5] [6] [7] [8] [9] . Traditionally, lower PEEP has been used [10, 11] , but low PEEP can lead to increases in oxygen desaturation and hypoxaemia [4, 12] and worsening of lung injury, indicated by greater use of rescue therapies and deaths after rescue therapy [9] . High PEEP can increase alveolar recruitment, but can decrease cardiac output and lead to further lung injury due to barotrauma and/or volutrauma [13] or overdistension [4, 9] . PEEP can be optimised to reduce hypoxaemia [6] and intrapulmonary shunting [7] and improve gas exchange [8] and oxygenation [4, 14, 15] , by maintaining recruitment of injured or collapsed alveoli [13] . In patients with ARDS, PEEP reduces ventilator-induced lung injury (VILI) [4, 9] , increases recruitment [14] [15] [16] , and reduces inflammatory mediators in plasma and bronchoalveolar lavage fluid [8] . MV strategies combining low tidal volumes with recruitment manoeuvres (RMs) and higher PEEP to prevent VILI have been hypothesized as ideal for lung protection [6, 17] . However, currently, there is still no standardized approach to the selection of this optimal PEEP, or to deciding how often PEEP should be adjusted or recalculated. Reports from experimental animal trials performed by Carvalho et al., Suarez-Sipmann et al. and Lambermont et al. [18] [19] [20] indicate that pigs induced with ARDS experience minimal respiratory elastance at a specific PEEP associated with higher oxygenation, maximum recruitment, and higher functional residual capacity, all without signs of lung overdistension. Equally, it has been proposed that PEEP should be set whereby the lung has minimal respiratory elastance (or maximum compliance), which could be clinically beneficial by balancing the risks of PEEP that is set too low or too high [21] [22] [23] . Aside from the work by Suter et al. [21] , Pintado et al. also showed that PEEP selection at minimal elastance is beneficial to patients [22] . Despite some consistent findings, the application of minimal elastance PEEP selection remains limited and hindered by the lack of an objective, reliable, and easy-to-use method to determine elastance at the bedside in real time. Chiew et al. showed the potential benefit of minimalelastance PEEP selection in a pilot study [23, 24] . Following the study, a phase-2 randomised controlled trial (RCT) was designed to assess mechanical ventilation at minimal elastance PEEP in patients with ARDS versus standard practice of care in a single-centre hospital. In particular, patientspecific respiratory system elastance and corresponding minimal elastance PEEP is determined using a validated model-based method and computer software [25] . This trial uses real-time-identified patient-specific respiratory system elastance, and thus the trial is named the Clinical utilisation of respiratory elastance (CURE) RCT. This manuscript presents the detailed clinical protocol for the phase-2 CURE RCT. This trial is registered with the Australian New Zealand Clinical Trial Registry (ANZCTR): ACTRN12613001006730. The CURE RCT is a two-arm RCT comparing modelbased mechanical ventilation (MBV) with current standard practice mechanical ventilation (SPV) in patients with a ratio of partial pressure of arterial blood oxygen (PaO2)/fraction of inspired oxygen (FiO2) (P/F ratio) ≤ 200. It is to be conducted in a single-centre hospital intensive care unit (ICU) at Christchurch Hospital in Christchurch, New Zealand. The primary objective is to assess the impact of model-based ventilation in PEEP selection (MBV) therapy on clinically significant patient outcomes and patient-centred quality of care metrics. The other objectives of this study include (1) to provide the knowledge and methods to make care more patient-specific and timely to optimise treatment and improve outcomes in a large cohort of critically ill patients and (2) to improve the understanding of the pathophysiological basis of critical illness through what we learn about the hourly and daily evolution of lung injury in terms of patient-specific elastance and response to care through this study. The primary outcome of this study is the area under the curve (AUC) of PaO2/FiO2 over the period of mechanical ventilation. Secondary outcomes include length of time of MV (LoMV), ventilator-free days (VFD) up to 28 days, ICU and hospital length of stay (LoS), AUC of SpO 2 (oxygen saturation)/FiO 2 during MV, number of desaturation events (frequency and fraction of time SpO 2 < 88%), changes in respiratory mechanics and chest x-ray index scores, rescue therapies (prone positioning, nitric oxide use, extracorporeal membrane oxygenation (ECMO)) and hospital and 90-day mortality. These outcomes and their corresponding four levels of specification based on Zarin et al., 2011 is shown in Table 1 . The secondary analysis includes comparison of the means of LoMV, VFD, hospital and ICU LoS, 90-day mortality, chest x-ray index scores and rescue therapies used. A difference in primary and secondary outcomes will show the impact of MBV compared to SPV. No difference would show that enhanced, model-based metrics of patient-specific condition have no effect on patientcentred or clinical outcomes. Either outcome will yield clinical guidance. Eligible patients are randomised to either the modelbased intervention group (MBV) or the control group (SPV). Both groups will have designated computer software to monitor their breathing [25] . The software uses real-time measurements of pressure and flow from the ventilator to objectively calculate the patient-specific and breath-specific respiratory system elastance for every breath [25] . Participants on MBV will undergo recruitment manoeuvres (RM), an initial maximum recruitment manoeuvre (RM Max ) or subsequent PEEP adjustment and monitoring procedure (PUMP) mini recruitment manoeuvres. The respiratory elastance at each PEEP step during these protocolised RMs is calculated and recorded. The software will recommend a patient-specific minimal-elastance PEEP to the clinicians in setting ventilator PEEP. Patients on SPV will have PEEP selected using current clinical practice without the aid of the software, but all breaths will be analysed and elastance recorded; clinical staff will be blinded to these data. Patients recruited into this study will be under constant supervision in the ICU. However, their outcomes will be measured based on the intention-to-treat principle, taking into account protocol variations, which naturally occur. These variations will be reported to the primary investigator at the earliest opportunity and followed up. There will be detailed training on the use of CURE equipment and on the protocol, to allow adherence to trial. This trial is based on the intention-to-treat principle. Thus, protocol amendments may be required to ensure patient safety and outcomes, and the primary investigators will instigate protocol amendments if necessary. The amendments will be reviewed by the data monitoring committee (DMC) to warrant patient safety and outcomes. The DMC may also refer protocol amendments based on outcomes of the interim analysis reports. Finally, if participant enrolment is slow, the protocol may be amended to allow faster recruitment. The trial involves critically ill participants who are mechanically ventilated. Thus, it is likely and acceptable for participants to be receiving medication related to any other concomitant comorbid conditions while participating in the CURE RCT. Participants in this study will not be concomitant to another study that would affect the results of this study. Participants will not be co-enrolled in another study that uses different oxygenation settings, recruitment manoeuvre procedures or anything that may affect the outcomes of this study. The following are the CURE RCT inclusion, exclusion and P/F ratio criteria. The inclusion criteria are: i. on any level of PEEP or FiO 2, or ii. P/F ratio ≤ 200 on FiO 2 = 50% and PEEP = 5 (see "P/F ratio criteria") The exclusion criteria are: The P/F ratio criteria are: 1. If the P/F ratio is ≤ 200, the patient is eligible for enrolment 2. If 200 < P/F ratio ≤ 300, set FiO 2 = 50% and PEEP = 5 cmH 2 O and repeat the analysis of arterial blood gases (ABG) within 10 min of the change to measure the new P/F ratio a. If the new P/F ratio is ≤ 200, the patient is eligible for enrolment b. If the new P/F ratio is > 200, the patient is not eligible for enrolment and, if appropriate, will be re-screened at a later time This trial will recruit patients who have a P/F ratio ≤ 200, a criterion in the definition of severe to moderate ARDS as defined by the ARDS Definition Task Force in the Berlin definition [26] . They will be eligible if their P/ F ratio is ≤ 200 on any level of PEEP and FiO 2 . Those patients with a 200 < P/F ratio ≤ 300 will be placed on PEEP = 5 cmH 2 O, and FiO 2 = 50%. If a subsequent P/F ratio is ≤ 200 they will also become eligible (see Additional file 1). The P/F ratio measured at FiO2 of 50% and PEEP of 5cmH 2 O is based on Villar et al. 2013 [27] . First, it is important to note that standard ventilation practice may include recruitment manoeuvres to increase lung recruitment and oxygenation. However, these clinical practices are widely variable and often not standardised. The recruitment techniques used to improve oxygenation and mechanics of ventilation in the intervention and control arms of this study are within the scope of standard ICU clinical practice. The protocols used will standardise these existing interventions to recruit lung volume and titrate PEEP. Study participants will be unable to consent to participation in this study prior to enrolment as they will be sedated and mechanically ventilated. It is also equally important to randomise participants to either arm of the RCT at the commencement of MV to ensure a fair comparison. Patients who have been ventilated ≤ 48 h are eligible for the CURE RCT. Given this time frame, the CURE RCT will recruit patients once family consent is obtained. However, if the treating clinician firmly believes a recruitment manoeuvre is in the best interests of the patient, and no family is available for consent, the participant will be enrolled and randomised and the appropriate protocolised recruitment manoeuvre will follow. In this case, delayed consent is obtained as early as possible. Once the participant recovers from their condition and is discharged from the ICU, we will seek their informed consent. In cases where the family cannot attend the hospital to sign a statement of assent, their opinion will be obtained by telephone in the first instance. Information about the study will either be made available by emailing them the information sheet and contacting them later by telephone, or the information sheet will be read to them over the telephone. The telephone conversation(s) and their opinions will be documented in the patient's medical record. As soon as the family is able to attend the hospital, they will be asked to sign the statement. If the family are not able to sign a statement during the patient's time in the ICU, they have the option of printing out the statement, signing it, and mailing, emailing or faxing it back. The sample study information and consent forms can be seen in Additional file 2. If the participant's family, relative or friend does not agree to their continued participation, they will be withdrawn from the study and we will seek agreement from them to use information related to mechanical ventilation that was collected up until that point. If a participant chooses to withdraw from the trial, we also will seek agreement to use information related to mechanical ventilation that was collected up until that point. If they do not agree, then all study information obtained will be destroyed. Participants will be block-randomised, with block sizes generated using a randomisation programme. The programme will randomly assign patients into either a control group or intervention group through a random block size (the block size is 4, 6, 8 or 10 patients). Eligible and consented patients will be block-randomised in a ratio of 1:1. No effort will be made to stratify the subgroups considered in the secondary analyses. By the nature of the intervention, CURE cannot be double-blinded. Un-blinding is not applicable due to the nature and setting of the intervention. All patient data collected are de-identified using a single patient numbering system. Patients are assigned to a study number to ensure no bias in the results. This system will be a simple incrementing scheme, such that patients randomised into the CURE trial are identified as study-001, study-002, etc. The V T is adjusted to 6-8 ml/kg per ideal body weight (IBW), and the maximum minute ventilation (V Emax ) to ≤ 0.2 L/kg/min. The IBW is measured using the patient's height and look-up table at the bedside or is calculated using the formulae: Women The DP is the plateau pressure (P Plat ) minus the PEEP. In patients with very severe ARDS the adjustment of V T to 6-8 ml/kg IBW may be injurious if the DP is higher than 15 cmH 2 O. A DP ≤ 15 cmH 2 O was associated with better patient survival when assessed using a multilevel mediation analysis of 3562 patients in nine RCTs of ARDS [28] . Therefore, the DP will be limited to ≤ 15 cmH 2 O at all times. In addition, during spontaneous ventilation, pressure support will be limited to ≤ 15 The ventilation rate is set between 12 and 20 breaths per minute. The aim is to keep the plateau pressure P plat ≤ 30 cmH 2 O. If necessary, V T may be reduced to as low as 4 ml/kg and the respiratory rate (RR) kept at ≤ 30 breaths per minute. CO 2 will frequently rise in severe lung injury (permissive hypercapnia) when patients are mechanically ventilated within these guidelines. However, if CO 2 is ≥ 80 mmHg or increased by ≥ 50% in the previous 4 h, the intensive care specialist on duty will be notified, and they may choose to deviate from these guidelines. All patients enrolled are to be ventilated using a pressurecontrolled mode, for example, the Bi-Level ventilation mode on the Puritan Bennett PB840 ventilator (Covidien, Boulder, CO, USA) or PC-SIMV+ on the Dräger Evita® In-finity® V500. Patients will be ventilated using Bi-Level/PC-SIMV+ mode, which allows unrestricted spontaneous breathing efforts to lessen ventilator dyssynchrony. However, during any recruitment manoeuvre procedures, synchronized intermittent mandatory ventilation (SIMV) with pressure-controlled (PC) ventilation is used and returned to original mode afterwards. Should patients already be ventilated using a ventilator incompatible with the CURE computer system, they will have their ventilator changed to a compatible ventilator for the trial. Patients will be transitioned to assisted spontaneous breathing (ASB) if they meet the weaning criteria (see "Weaning"). In severe ventilator dyssynchrony, a very high respiratory drive may result in sub-atmospheric circuit pressures and risk of aspiration of gastric contents around the endotracheal cuff. If a participant has a high respiratory drive on Bi-Level/PC-SIMV+ ventilation, producing a fall in airway pressure during inspiration, muscle relaxants will be considered to facilitate controlled breathing. However, if the clinician feels the participant may benefit from breathing spontaneously, transition to ASB may be made if they substantially meet the weaning criteria (see "Weaning"). However, spontaneous breathing efforts may mask high trans-pleural pressures and produce high levels of regional lung strain. Oesophageal pressure will not be measured during this trial. If the treating clinician is concerned about patient self-inflicted lung injury (P-SILI) [29, 30] , they will consider using muscle relaxants to control ventilation. Patients will not undergo any procedures using a cough-assist machine prior to weaning and transitioning to spontaneous breathing. However, the treating clinician may use a cough-assist machine to aid secretion removal (during spontaneous breathing) if they believe it is in the patient's best interests. Finally, in any circumstances where the patient is planned to be temporarily disconnected from the ventilator, their endotracheal tube will be clamped to prevent de-recruitment. To ensure a fair comparison, all CURE study participants will have inspired oxygen levels titrated to achieve the following pulse oximetry saturations: i. SpO 2 = 93-95% if FiO 2 is less than 60% ii. SpO 2 = 90-92% if FiO 2 is greater than or equal to 60% The aim is to spend greater than or equal to 90% of time in the target range. The FiO 2 should only be increased above 21% if these targets are not met, using 5% increments starting with a FiO 2 = 25%. There is natural variability in SpO 2 . To avoid toggling between two FiO 2 levels, 10 min of settling time will be allowed before changing the FiO 2 . The best FiO 2 is chosen to keep the saturation within the specified targets ranges over 90% of the time. Patients are kept at 30°head up whenever possible. This position maximises recruitment of the lung and may reduce the risk of aspiration. Wherever possible, patients should be rolled from supine to right-side down, back to supine, then to left-side down. This turning of patients is ideally performed every 3 h. Transient hypoxaemia frequently occurs after a patient has been turned and may be worse if there is inadequate PEEP. Hypoxaemia may also become more severe if participants are rolled from left-side down to right-side down due to cyclical de-recruitment of the non-dependent lung and re-recruitment of the dependent lung. This cyclical de-recruitment of the lung has the potential to contribute to VILI. Thus, patients with severe lung injury may be very intolerant of being turned. In some instances, the lungs may need to be re-recruited. If desaturation does occur, this will be recorded as a serious adverse event (SAE). Prone positioning of patients may be considered if the P/F ratio is ≤ 100 and FiO 2 ≥ 60%. Patients randomised to the intervention arm (MBV) may still undergo a protocolised recruitment manoeuvre. For patients in the standard practice ventilation arm (SPV) a staircase recruitment manoeuvre is left to clinical judgement. The primary outcome of this trial is the AUC of the P/F ratio. For this reason, mandatory daily ABG recordings are performed for up to 10 days after enrolment. ABGs are taken around 0600 hours and 1800 hours. The ABGs are also acquired within 60 min of any recruitment manoeuvre procedure and 30-60 min after the recruitment manoeuvre procedure. The added ABGs from RM procedures will be used in secondary analysis, but will be omitted during primary analysis to ensure the same number of data points per day for all patients. Patients randomised to the model-based ventilation (MBV) cohort will remain in the protocol up to 10 days. Thereafter, they will receive the same care as participants assigned to standard practice ventilation (SPV). However, if participants have been extubated, but then require intubation and re-ventilation at any time within 10 days of enrolment, they will return to the original assigned protocol (MBV or SPV). All patients will receive standard care beyond 10 days of enrolment, and their data, including ABG recordings, will continue to be collected for up to 28 days. This intervention schedule is shown in Fig. 1 . Procedures for the control group (SPV) The procedures are as follows: 1. PEEP is selected as per standard practice. 2. The decision to carry out a staircase recruitment manoeuvre will be based on clinical judgement. The protocol for performing the staircase recruitment manoeuvre is explained in "Recruitment manoeuvres (RM)". 3. ABGs will be taken twice daily. 4. Ventilator data are collected continuously until the ventilator is disconnected. The procedures are as follows: 1. For patients included in the MBV (intervention) group, the PEEP and MV will be guided by clinicians using bedside computers, while maintaining V T and FiO 2 . iii. When they have a new neurological condition. iv. They are awake and breathing normally without evidence of respiratory distress, and where sedation (with or without paralysis) is not considered to be in their best interests. 4. ABGs will be recorded twice daily and before and after any recruitment procedure. 5. Data will be collected continuously until the patient is disconnected from the ventilator. Patients enrolled in this study will undergo RMs. The RMs are only carried out by senior medical staff or senior trainees familiar with this technique. RM Max and PUMPs are for participants randomised to the MBV protocol arm only. Patients assigned to the SPV arm may undergo a staircase RM (SRM) at the discretion of the treating clinician according to standard practice. (see "Standard practice staircase recruitment manoeuvre"). All RMs will be performed in SIMV pressure controlled (PC) ventilation mode. The peak inspiratory pressure (Pi) is set to achieve a V T of 6-8 ml/kg IBW. Preferably, V T should result in DP ≤ 15 cmH 2 O above PEEP. Before and after each RM, ABGs (pH, PaCO 2 , PaO 2 , HCO 3 ), SpO 2 , end-tidal CO 2 partial pressure (ETCO 2 ), FiO 2 , PEEP, RR), and V T will be recorded. In addition, during the RM max (model-based ventilation arm) or (standard practice ventilation (SRM) arm, at each PEEP increment, Heart rate, rhythm, mean arterial pressure (MAP), SpO 2 , FiO 2 , V T , RR, ETCO 2 and rates of use of vasoactive drugs will be recorded. This data will be valuable in assessing the safety of the RM Max , PUMP and SRM. In many cases, where the lung stiffness, or respiratory elastance (E) is high, it will not be possible to deliver a V T of 6 ml/kg IBW. Furthermore, during the RM, the delivered V T may fall further as the elastance increases. As a result, it may be necessary to increase the respiratory rate to accommodate the reduction in minute ventilation. For example, if E is > 40 cmH 2 O/L (or compliance < 25 ml/cmH 2 O) in a patient weighing 58 kg (IBW), (normal range 15-20 cmH 2 O/L), the V T will be < 6 ml/kg (< 350 ml) when the driving pressure is 15 cmH 2 O. It is important that oxygenation targets in both arms are carefully followed to ensure a fair comparison between them. The SpO 2 will be kept in the target range prior to any RM. This approach allows small decreases in oxygenation to be detected during the decremental PEEP phase of the recruitment manoeuvre, while also providing a sufficient buffer in the event of significant de-saturation due to ventilation perfusion (V/Q) mismatch. V/Q mismatch increases with higher airway pressures when pulmonary arterial blood is shunted away from the pulmonary capillaries by-passing aerated regions of the lung. Before performing any RM, the following criteria are considered. Any RM must be delayed until these conditions are corrected in the consideration of the following at-risk patient conditions: 1. Haemodynamic instability (e.g. ongoing haemorrhage). 2. Not optimally resuscitated with fluids (e.g. stable blood pressure, but pulse pressure variation ≥ 12% because of inadequate left ventricular preload)? This is only applicable in the absence of spontaneous breathing. 3. Evidence of barotrauma since enrolment? a. If there is new barotrauma, RMs must not be attempted and the participant will be withdrawn from trial. They will continue to be observed and followed up. A SAE will be reported. Once the RM checklist conditions are met, the patient can be prepared for a RM by ensuring: RMs should be terminated if at any time during the RM any of the following changes persist for more than 3 min: 1. Desaturation, with SpO 2 < 88%. 2. New bradycardia (heart rate < 60 beats per minute) or, 3. New tachycardia (heart > 140 beats per minute) or, 4. New arrhythmia leading to new bradycardia (2) or new tachycardia (3) or, 5. New hypotension (reduction in MAP by 40% or MAP < 60 mmHg). This RM termination criteria applies to all RM procedures in both arms. The RM Max is a computer-guided staircase RM procedure in the MBV intervention arm. This method is designed to safely increase the inspiratory pressure to a maximum airway pressure of 40-43 cmH 2 O, with DP limited to 15 cmH 2 O, and maximum PEEP limited to 25-28 cmH 2 O. The RM Max is guided by the CURE software using a validated model-based method, which estimates elastance to determine the optimal PEEP [23, 31] . The RM Max is carried out by intensive care specialists or senior trainees familiar with this technique. This procedure is only carried out during working hours (0800-1800 hours), but preferably within 4-6 h of enrolment. However, for patients enrolled overnight, unless there are compelling reasons to carry out an RM Max , this procedure may be delayed till the following morning (0800 hours). Contra-indicated preconditions to an RM Max are excluded using the RM checklist. If it is safe to proceed, the patient is prepared for the RM Max . The following instructions are given to the clinician: The RM Max may be repeated only when the following conditions are met: 1. If there is a significant change in the participant's condition, e.g. new severe hypoxaemia (SpO 2 < 90% and FiO 2 ≥ 60%; P/F~100) and 2. Patient conditions for which lung recruitment is contraindicated are excluded (e.g. endobronchial intubation, mucous plugging, pneumothorax etc) and 3. Analgesia and sedation and patient position have been optimised (consider small changes to respiratory rate, V T and pressure support, or a rocuronium infusion) and 4. The PUMP fails to improve oxygenation PUMP: PEEP adjustment and monitoring procedure PUMP is a regular mini-recruitment manoeuvre procedure designed to adjust PEEP based on patient-specific changes in condition. This mini-RM is also guided by CURE software and moves between ±4 cmH 2 O from the current PEEP. PUMP should be performed twice daily during normal working hours (0800-1800 hours) or at any other time if lung de-recruitment is considered to be the likely cause of new desaturation. The Pi will be left the same as in the current ventilator settings. To ensure a PUMP can be safely carried out, the RM checklist and preparation steps are to be followed. If the checklist preconditions are met, the PUMP may be carried out. The following instructions are given to the clinician: 1. Titrate sedation so the patient is not verbally responsive and has loss of their eyelash reflex. Use fentanyl or morphine increments with propofol to provide a "balanced" deeper sedation level. Give rocuronium 0. If the oxygenation does not improve with the aforementioned interventions, then PEEP may be increased in increments of 2 cmH 2 O. If the PEEP is ≥ 15 cmH 2 O and FiO 2 is ≥ 60%, (P/F~100) in spite of addressing the aforementioned points, a staircase recruitment manoeuvre (SRM) may be considered if the clinician feels this is in the best interests of the patient. The SRM procedure does not utilise the CURE software to perform recruitment and therefore the software will not guide the user, nor make any PEEP suggestions. The software will still record airway pressure and flow through this procedure. To ensure a SRM, can be safely carried out, the RM checklist and preparation steps are to be followed. If the checklist preconditions are met, the SRM may be carried out (note, the SRM procedure does not utilise CURE software to perform recruitment). The following instructions are given to the clinician: 1. Titrate sedation so the patient is not verbally responsive and has loss of their eyelash reflex. Use fentanyl or morphine increments with propofol to provide a "balanced" deeper sedation level. Ventilator dyssynchrony occurs when a patient's spontaneous respiratory efforts are not synchronised with the ventilator. This commonly causes agitation and respiratory distress; often described as "fighting the ventilator". Dyssynchrony should be considered in patients with increased respiratory efforts, unexplained agitation, tachycardia, or sweating. Ventilator wave forms can be used to identify dyssynchrony. In participants assigned to MBV, dyssynchrony will often cause large spikes in the elastance recordings. The CURE soft algorithm does not account for patient breathing efforts and "sees" inspiratory effort as a rapid reduction in lung elastance [32, 33] . In contrast, coughing, breath-holding, and other dyssynchronous efforts may cause an apparent increase in elastance [34] . Figure 2 shows an example of ventilator dyssynchrony in a pressure-controlled mode. Dyssynchrony may be seen as negative deflections ("M" waves) in the flow-time waveform, as shown in Fig. 2 . In contrast the airway pressure may only be changed minimally by patient effort. It is important to exclude reversible mechanical causes that might lead to patient distress and ventilator dyssynchrony. Endobronchial intubation, obstruction of a major bronchus or pneumothorax should be excluded. Usually ventilator dyssynchrony can be managed by increasing sedation. However, in many cases it may be preferable to use intermittent muscle relaxants to fully control ventilation. It also may be helpful to trial the patient on assisted spontaneous breathing (ASB) to improve ventilation synchrony, if PEEP is ≤ 10 cmH 2 O and the FiO 2 is ≤ 40%. However, caution should be exercised, lest the patient become exhausted (see "Weaning"). These guidelines are a pragmatic and consistent way to transition patients to ASB. Weaning is challenging and difficult to protocolise because there are many different factors to consider. For this reason, the weaning process is typically determined by clinical judgement. However, the guidelines presented here are set to ensure consistency of care. ASB is considered when the participant's condition is improving. They should preferably be afebrile, have resolution of the underlying processes that led to their ICU admission and improving gas exchange. They should have improving muscle strength, decreasing sedation requirements with an improving Glasgow Coma Score (GCS) and Richmond Agitation Sedation Score (RASS) between −3 and +1. Generally, the FiO 2 should be ≤ 40% and PEEP ≤ 10 cmH 2 O. If the following are substantially present, then participants may be transitioned to ASB: 1. Improving condition 2. Minute ventilation acceptable (VE) ≤ 0.2 L/kg 3. FiO 2 ≤ 40% 4. SpO 2 93-95% 5. pH ≥ 7.3 6. Heart rate ≤ 120 beats/min 7. Low vasoactive drug requirements (noradrenaline + adrenaline ≤ 10 mcg/min) The following instructions or recommendations are used to guide transition to ASB: The following observations should be made: oxygenation, re-sedate and revert back to the previous controlled ventilation mode (Bi-Level or PC-SIMV+ or equivalent); e. If the oxygenation has not improved after 12 h on controlled ventilation, or there is an unanticipated new problem causing deterioration, the participant should return to their previously assigned ventilation arm (MBV or SPV), e.g.: iv. New lung injury/de-recruitment/aspiration/ sepsis. v. Haemodynamic instability. vi. Need to return to the operating room or to undergo invasive procedure. 2. If there is continual improvement, proceed towards separation from mechanical ventilation (extubation or continuous positive airway pressure (CPAP) via a tracheostomy). A simplified flow chart of the patient enrolment process can be found in Additional file 1. All study data, including ventilation data, patient and family/friend consent, SAE reports and other documentation will be stored in a repository. Consolidated standards of reporting trials (CONSORT) Figure 3 shows the CONSORT diagram for the CURE RCT. Patients recruited into the CURE RCT will have the following data collected. Patient demographic and history The following data will be collected: 1. Patient gender, height, weight and ethnicity 2. Primary patient diagnosis contributing to ARDS or impaired lung function 3. Secondary patient diagnosis contributing to ARDS or impaired lung function 4. Relevant past medical history, e.g. smoking, medication, cardiovascular disease 5. Chest x-ray score derived by the Murray Index [35] Patient mechanical ventilation data Data on patient airway pressure and flow generated from the mechanical ventilator will be recorded using the CURE software (CURE Soft.) provided with the RCT. The patient data are backed up regularly to external storage with encryption applied using VeraCrypt encryption software [36] . The following information will be collected: or ventilation-free days 9. Amount of sedation -to account for possible data variation resulting from use of different sedatives 10. Duration of ICU stay 11. Frequency and duration of renal support therapies 12. All causes of ICU, hospital and 90-day mortality No person or authority will have access to the participant's blood. The blood samples are not stored. They are discarded and incinerated as soon as practicable, in accordance with NZS 4304:2002 "Healthcare Waste Management". All CURE RCT data will be stored at the University of Canterbury (UC). All paper forms (patient sheets, consent forms, etc.) will be scanned and stored at the UC. All electronic data will be stored in double-encrypted repository and only the participating researchers have access to it. Currently there are no plans for sharing the data, but if requested, data may be shared. Participants in the study can request their copy of data. The data will be backed up weekly and again, once the participant has finished the trial and left the hospital. This task will only be performed by the participating researchers. Any protocol variations will be followed up and noted. The CURE RCT will store data for 20 years. The trial will utilise a primary composite end point incorporating the AUC of the P/F ratio over the period of MV. Every participant in the intervention (MBV) group is compared with every participant in the control (SPV) group. Test statistics will be calculated using the onesided Wilcoxon rank sum test at alpha of 0.05. If results show no statistically significant difference between the intervention and control, it will result in the rejection of the intervention treatment as a standard of care and thus, the secondary clinical outcome assessments will include the number of desaturation events measured as peripheral capillary oxygen saturation < 90%, LoMV, VFD for 28 days, the quality of mechanical ventilation care measured as the AUC of SpO 2 /FiO 2 and chest x-ray index scores over time. The test statistic will be calculated using the one-sided Wilcoxon rank sum test at alpha of 0.05. A difference in the primary outcome will show the impact of MBV compared to SPV. No difference would show that enhanced, model-based metrics of the patient-specific condition have no effect on patient-centred or clinical outcomes. Either outcome will yield clinical guidance. A Monte-Carlo simulation was performed to determine the sample size and determined that a minimum effective sample size of approximately 160 patients per arm is required to identify a 25% reduction in median LoMV, with 0.8 power, at a double-sided significance level of 5% [37] . A linear alpha spending approach will be used for early termination of the trial for safety. Linear alpha spending falls between a Pocok and O'Brien-Fleming boundary [38] . With analysis points of 50, 75, 100, 125 and 160 patients per arm, and assumed control group mortality of 0.2, the mortality difference required to stop the trial (Mortality Intervention -Mortality control ) at each analysis point respectively is 0.20, 0.16, 0.14, 0.12 and 0.10. This approach has cumulative α = 0.025. Ethics approval has been filed with the New Zealand National Health and Disability Ethics Committee. The CURE RCT clinical protocol and data usage has been filed with the New Zealand South Regional Ethics Committee (reference number 14/STH/132). The CURE trial is also registered in the Australian New Zealand Clinical Trial Registry (ACTRN12614001069640). All results and any subsequent analysis will be published and only the participating investigators will be authors. Currently there is no plan to share data with other organisations. The data collected in this study will also be used for future research. Adverse event (AE) and SAE reporting (AEs are defined as any unexpected change in physiology in a study participant associated with either the RM max or PUMP. This does not necessarily have to have a causal relationship with the aforementioned procedures. Typically, this would be an unexpected, non-life-threatening event, which rapidly resolves following simple corrective measures. For example, hypotension will occur in most participants undergoing an RMmax or PUMP. However, if the procedure had to be shortened or abandoned, but the participant recovered with simple corrective measures (e.g. temporarily increasing noradrenaline by ≥ 5 mcg/min) or giving > 500 ml fluid bolus) this would be recorded as an AE. It is very important these events are accurately recorded as risk factors for AEs that need to be defined when carrying out RMs (as seen in Additional file 3). SAEs are defined as any untoward medical occurrence that (1) results in death; (2) is life-threatening; (3) prolongs hospitalisation or (4) results in disability or incapability. However, baseline mortality in the patients in intensive care who are enrolled in the trial will likely be high due to the critical illness necessitating admission to the ICU. Despite attempts at prevention, trial participants will frequently develop lifethreatening organ failure(s) unrelated to study interventions. Events that are a part of the natural history of the primary disease process or expected complications of critical illness will not be reported as SAEs in this trial. Additionally, events already defined and reported as study outcomes, such as mortality or readmission to the ICU, will not be labelled and reported separately as SAEs unless they are considered to be causally related to the study intervention or are otherwise of concern in the investigator's judgement. SAEs will be reported to the principal investigator within 24 h of any investigator becoming aware of the event. The minimum information to report includes: The nature of the event The time the event commenced and ceased An investigator's opinion of the relationship between study involvement and the event (not related, unlikely, possibly, probably or definitely related) Whether treatment was required for the event and what treatment was administered SAEs could include pneumothorax, hypotension leading to cardiac arrest, transient desaturation leading to severe or prolonged desaturation, tachycardia, bradycardia, arrhythmia, anaphylaxis and unintended protocol deviations. In the unlikely event of a physical injury to the participant as a result of their participation in this study, they will be eligible to apply for compensation from the Accident Compensation Corporation (ACC) NZ within its limitations. If the participant's family/friend have any questions about the ACC, they will be able to ask the researchers for more information before they agree to take part in this trial. ACC cover is not automatic and their case will need to be assessed by the ACC according to the provisions of the 2001 Injury Prevention Rehabilitation and Compensation Act. If the claim is accepted by the ACC, the patient still might not receive compensation. This depends on a number of factors such as whether they are earners or non-earners. The ACC usually provides only partial reimbursement of costs and expenses and there may be no lump sum compensation payable. There is no cover for mental injury unless it is a result of physical injury. If your relative or friend has ACC cover, generally this will affect their right to sue the investigators. An independent DMC comprising experts in clinical trials, biostatistics and intensive care medicine is established before patient enrolment, to review all trial protocols and oversee and advise on this trial. The DMC will be forwarded a copy of all SAE reports as soon as they become available to the trial investigators. The DMC will review all SAE reports that they receive and report back to investigators if any further action is required. The steering committee of the CURE RCT comprises the primary investigators Geoff Shaw, Geoff Chase, Chris Pretty and Yeong Shiong Chiew. The clinical data are collected by research nurses in the ICU and mechanical ventilation data and oxygenation (bedside monitor) data will be collected by researchers from the UC. All study data will be stored in the double-encrypted repository at UC. Data will be interpreted by participating researchers. The open and closed case interim reports will be compiled by Paul Docherty every 6 months and when 50 and 100 patients have been included. The DMC will have authority to continue or stop of the trial based on the interim reports. Mechanical ventilation using PEEP set at minimum elastance has long been investigated in both experimental and clinical trials. These studies ranged from healthy patients under general anaesthesia to those with ARDS. However, only a few studies have investigated the clinical potential of PEEP set at minimum elastance. Recent studies by Pintado et al. [22] and Chiew et al. [23] have shown the potential and feasibility of ventilating patients using minimum elastance PEEP. However, setting PEEP based on elastance is problematic due to the increased need of muscle relaxants, protocol burden and potential contradictory findings [22, 23] . The pilot trial was also underpowered and thus, a larger clinical trial such as the CURE RCT is required to provide further insight and validate the potential benefit of optimising mechanical ventilation PEEP with model-based methods. The CURE RCT implements a protocolised staircase PEEP recruitment manoeuvre together with novel computer software to calculate respiratory system elastance in real time. The computer software, CURE Soft [25] , uses a single compartment lung model [39] together with other model-based approach [24, 40] to aid clinicians during PEEP selection. This process potentially reduces selected PEEP variability and provides more consistent clinical guidance. There are several limitations of the CURE RCT design that should be noted. In particular, the RM is a doublestaircase manoeuvre and is design specific. Studies have shown that not all patients benefit from RMs [41, 42] , and the benefit of an RM is dependent on the patientspecific disease state, as well as the design of the RM. The double-staircase RM in this trial was designed to assess lung recruitment and provide consistent PEEP titration. It is noted that not all patients included in this study will necessarily demonstrate alveolar lung recruitment. Another limitation worth noting is the control group clinical protocol. Clinically, there is relatively little consensus an optimal mechanical ventilation mode. Thus, the standard practice ventilation in the participating hospital relies on general approaches [2] and is highly variable between clinicians. There may be no equal comparator for a mechanical ventilation study resulting from this variability. Patients recruited to the CURE RCT will have the MV mode set to bi-level ventilation in both the intervention and the control group; it is debatable that bi-level ventilation may lack certain ventilation advantages. However, this procedure will reduce variability and provides the opportunity for fair comparison between groups. In the participating ICU, there are > 700 patients per year requiring invasive MV; however, only an average of < 5 patients were diagnosed with ARDS as the primary diagnosis per year. ARDS is nearly always regarded as a complication of an acute process. One concern is that the desired sample size cannot be achieved. However, this number is also too small compared to reports [43, 44] . The small number may be due to changes in the ARDS definition [27, 45] and misdiagnosis [46] . Estenssoro et al. [46] report that misdiagnosis could occur due to delayed screening. Thus, in the CURE RCT, any patient requiring invasive mechanical ventilation is screened immediately, as per Villar et al. [27] , whereby the P/F ratio is measured at PEEP = 5 cmH 2 O, and FiO 2 = 50%. Equally, retrospective screening was also performed and identified > 200 patients eligible for the trial per year. Hence, a 3-year study is planned to achieve the target sample size at an estimated recruitment rate of 50%. Optimising patient-specific mechanical ventilator settings remains a huge clinical challenge due to patient disease variability, as well as clinical practice variability. Thus, there is a need for a method to provide consistent patient-specific treatment. The CURE RCT is the first single-centre large clinical RCT using model-based minimum elastance PEEP selection in mechanical ventilation. It provides a means to select patient-specific PEEP in a consistent fashion and patient outcomes are compared to current practice. The CURE RCT investigation group hope that the results from this trial will support the use of model-based methods to estimate optimal PEEP, and will serve as a platform to assess other patient-centred outcomes in future mechanical ventilation studies in ARDS/ALI. analysis. TD, BB, AS and KM provided advice on the study protocol. All authors read and approved the final manuscript. This research is funded by the New Zealand Health Research Council (HRC) (HRC reference 13/213) under project PulMoDS: Pulmonary Model-based Decision Support to Optimise ARDS/ALI Care. HRC does not have any role in this study and future analysis. They were not part of study design and will not be part of any management or analysis of the trial. Currently there are no plans to share any of the data. The CURE RCT clinical protocol and data usage has been granted by the New Zealand South Regional Ethics Committee (reference number 14/STH/ 132). The CURE RCT is registered in the Australian New Zealand Clinical Trial Registry (ACTRN12614001069640). The trial was registered on 22 September 2014 (https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=36683 8&isReview=true). The CURE RCT will recruit patients once family consent is obtained. However, if the treating clinician firmly believes a recruitment manoeuvre is in the best interests of the patient, and no family is available for consent, the participant will be enrolled and randomised and the appropriate protocolised recruitment manoeuvre will follow. In this case, delayed consent is obtained as early as possible. Once the participant recovers from their condition and is discharged from the ICU, we will seek their informed consent. Not applicable. Mechanical ventilation in ARDS: a state-of-the-art review Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome Effect of a protective-ventilation strategy on mortality in the acute respiratory distress syndrome Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome A high positive end-expiratory pressure, low tidal volume ventilatory strategy improves outcome in persistent acute respiratory distress syndrome: a randomized, controlled trial Ventilation strategy using low tidal volumes, recruitment maneuvers, and high positive end-expiratory pressure for acute lung injury and acute respiratory distress syndrome: a randomized controlled trial Positive end-expiratory pressure setting in adults with acute lung injury High levels of PEEP may improve survival in acute respiratory distress syndrome: a meta-analysis Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome Low mortality associated with low volume pressure limited ventilation with permissive hypercapnia in severe adult respiratory distress syndrome Positive end-expiratory pressure The preventive role of higher PEEP in treating severely hypoxemic ARDS Combined effects of ventilation mode and positive end-expiratory pressure on mechanics, gas exchange and the epithelium in mice with acute lung injury Reversibility of lung collapse and hypoxemia in early acute respiratory distress syndrome How large is the lung recruitability in early acute respiratory distress syndrome: a prospective case series of patients monitored by computed tomography Computed tomography assessment of positive end-expiratory pressure-induced alveolar recruitment in patients with acute respiratory distress syndrome Ventilation and weaning practices in Australia and New Zealand Positive end-expiratory pressure at minimal respiratory elastance represents the best compromise between mechanical stress and lung aeration in oleic acid induced lung injury Use of dynamic compliance for open lung positive end-expiratory pressure titration in an experimental study Comparison of functional residual capacity and static compliance of the respiratory system during a positive end-expiratory pressure (PEEP) ramp procedure in an experimental model of acute respiratory distress syndrome Optimum end-expiratory airway pressure in patients with acute pulmonary failure Individualized PEEP setting in subjects with ARDS: a randomized controlled pilot study Feasibility of titrating PEEP to minimum elastance for mechanically ventilated patients Model-based PEEP optimisation in mechanical ventilation The clinical utilisation of respiratory elastance software (CURE Soft): a bedside software for real-time respiratory mechanics monitoring and mechanical ventilation management The ARDS Definition Task Force, Definition TB A universal definition of ARDS: the PaO2/FiO2 ratio under a standard ventilatory setting-a prospective, multicenter validation study Driving pressure and survival in the acute respiratory distress syndrome Mechanical ventilation to minimize progression of lung injury in acute respiratory failure Ventilation-induced lung injury exists in spontaneously breathing patients with acute respiratory failure: Yes A minimal algorithm for a minimal recruitment model-model estimation of alveoli opening pressure of an acute respiratory distress syndrome (ARDS) lung Patient-ventilator asynchrony during assisted mechanical ventilation Patient-ventilator asynchronies: may the respiratory mechanics play a role? Observational study of patientventilator asynchrony and relationship to sedation level ☆ An expanded definition of the adult respiratory distress syndrome effective sample size estimation for a mechanical ventilation trial through monte-carlo simulation: length of mechanical ventilation and ventilator free days Designs for group sequential tests Lung mechanics at the bedside: make it simple Time-varying respiratory system elastance: a physiological model for patients who are spontaneously breathing Recruitment maneuvers for acute lung injury New and conventional strategies for lung recruitment in acute respiratory distress syndrome The Australian and New Zealand Intensive Care Society Clinical Trials Group. Incidence and mortality of acute lung injury and the acute respiratory distress syndrome in three Australian States Incidence and outcomes of acute lung injury Report of the American-European consensus conference on ARDS: definitions, mechanisms, relevant outcomes and clinical trial coordination Impact of positive end-expiratory pressure on the definition of acute respiratory distress syndrome Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations The authors wish to thank the Christchurch Hospital intensive care medical, nursing and technical staff for the support of this clinical trial. The authors declare that they have no competing interests. The trial has not started recruiting yet. The trial is estimated to start in December 2019 in Christchurch Hospital Intensive Care Unit. The trial is estimated to be completed by May 2022. This is protocol version number 2.0, dated 20 August 2019. Supplementary information accompanies this paper at https://doi.org/10. 1186/s13063-019-4035-7.Additional file 1. Screening process diagram.